T-ALL cells in the CNS are dependent on support from tumor-associated myeloid cells for survival. (A) Schematic diagram of in vitro coculture assays to assess the ability of myeloid cells from the CNS to promote survival of CNS-resident T-ALL cells. (B-C) Quantification of viable T-ALL cells isolated from the spleen (B) or CNS (C), assessed by flow cytometry 6 to 7 days after the initiation of cell culture in the presence or absence of leukemia-associated myeloid cells from the same organs. Viable cell numbers were normalized to wells in which “T-ALL” was cultured alone; circles represent the average of triplicate wells per experiment, with distinct primary LMO2 T-ALL stocks denoted via distinct colors. Bars represent mean + standard error of the mean (SEM) of data from 12 compiled experiments. The 2 graphs on the right show the same data as in the left graph but separated into Hhexhigh (left) and Hhexlow (right) leukemia subtypes. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001 (using the unpaired Student t test [Mann-Whitney U] in panels B-C).